Multi-domain Aspect Extraction Based on Deep and Lifelong Learning

Dionis López, Leticia Arco

Research output: Chapter in Book/Report/Conference proceedingChapterResearchpeer-review


Opinions concerning features or aspects of people, entities, products or services are some of the most important textual information. Several methods try to solve the aspect extraction task needed in sentiment analysis by using Deep Learning techniques in specific domains. However, catastrophic forgetting appears when these methods are used to learn aspects of multi-domains. In this paper, we propose a new approach to achieve aspect extraction in multi-domains based on Deep and Lifelong Learning techniques. Our proposal reduces catastrophic forgetting and improves one of the principal state-of-the-art results.
Original languageEnglish
Title of host publicationProgress in Pattern Recognition, Image Analysis, Computer Vision, and Applications
EditorsIngela Nyström, Yanio Hernández Heredia, Vladimir Milián Núñez
PublisherSpringer Verlag
Number of pages10
ISBN (Electronic)978-3-030-33904-3
ISBN (Print)978-3-030-33903-6
Publication statusPublished - 22 Oct 2019

Publication series

NameLecture Notes in Computer Science book series
PublisherSpringer, Cham


  • opinion mining
  • aspect extraction
  • deep learning
  • lifelong learning


Dive into the research topics of 'Multi-domain Aspect Extraction Based on Deep and Lifelong Learning'. Together they form a unique fingerprint.

Cite this